Stages and Exit Criteria

Why stage definitions matter

A pipeline is only useful if a stage means the same thing to everyone. Where stages are defined by the seller's activity — "proposal sent", "demo done" — the pipeline records what the seller did rather than where the buyer is, and it will always be optimistic.

The fix is to define stages by observable buyer behaviour and to attach exit criteria: specific, verifiable conditions that must be true before an opportunity can move forward.

A workable stage model

Stage	Exit criteria — verifiable facts
Qualified lead	Contact made, a relevant problem acknowledged, fit plausible
Discovery	Problem and consequences quantified, decision process understood, budget range established, next meeting agreed
Solution agreed	Buyer confirms the proposed approach addresses the problem; success criteria agreed
Proposal / business case	Proposal issued against agreed criteria; economic buyer identified and engaged; internal approval route known
Negotiation	Buyer has indicated intent to proceed subject to terms; commercial discussion under way
Verbal / pending signature	Decision made; remaining steps are contractual with named owners and dates
Closed won / lost	Signed, or a documented reason for loss

The discipline is simple and hard: an opportunity cannot advance until the criteria are met, regardless of how positive it feels. Most forecast inaccuracy is caused by deals sitting in a late stage they never earned.

Two rules that keep a pipeline honest

• Every open opportunity has a next step with a date. No next step means it is not active, and should be reverted or closed.
• Close dates change only for a stated reason. A date that slips silently each month is a deal that is not real, and the pipeline should show that.

Stage-weighted probability

Assigning a probability to each stage allows a weighted pipeline value. It is useful in aggregate across many deals and misleading on any individual one — a deal is won or lost, not sixty per cent won. Probabilities should be derived from your own historical conversion rates by stage, not invented, and they should be reviewed as those rates change.